Drift from field crop sprayers: The influence of spray application technology determined using indirect and direct drift assessment means

Society’s preoccupation with the use of plant protection products has increased significantly during the last few years. When pesticides are applied some of the spray may move beyond the intended area to undesirable areas which might have consequences such as damage to sensitive adjoining crops, water contamination, health risks to animals and people and a lower dose than intended on the target field.

The general objective of this research was to investigate the influence of spray application technology on the amount of spray drift from field crop sprayers. Therefore, indirect and direct drift assessment means were used and compared namely PDPA laser, wind tunnel and field drift measurements. A reference spraying was used for a comparative assessment of the different evaluated spray application techniques. This reference spraying was defined as a standard horizontal boom sprayer with a spray boom height and nozzle distance of 0.50 m, ISO 03 standard flat fan nozzles at a pressure of 3.0 bar and a driving speed of 8 km.h^-1. Besides this reference spraying, different other spray application techniques were tested to assess the effect of nozzle type (standard flat fan, low-drift, air inclusion), nozzle size (ISO 02, 03, 04 and 06), spray pressure (2.0, 3.0 and 4.0 bar), driving speed (4, 6, 8 and 10 km.h^-1), spray boom height (0.30, 0.50 and 0.75 m) and air assistance.

The developed PDPA laser measuring set-up is composed of a spray unit, a three-dimensional automated positioning system and an Aerometrics PDPA laser system in a controlled climate room which measures droplet size and velocity characteristics using a well-defined measuring protocol.
Wind tunnel measurements, performed in Silsoe Research Institute (SRI), were used to measure airborne and fallout spray volumes under directly comparable and repeatable conditions for single and static nozzles. Based on these measurements, drift potential reduction percentages (DPRP), expressing the percentage reduction of the drift potential compared with the reference spraying, were calculated following three approaches. The first approach was based on the calculation of the first moment of the airborne spray profile (DPRP_V1). In the second and third approach, the surface under the measured airborne (DPRP_V2) and fallout (DPRP_H) deposit curve were used.
For the field measurements, sedimenting spray drift was determined by sampling in a downwind area at 24 different positions using horizontal drift collectors in combination with a fluorescent tracer with measurements up to 20 m from the directly sprayed zone. Meteorological conditions were continuously monitored.
Based on 27 drift experiments with the reference spraying at various environmental conditions the important effect of atmospheric conditions on the amount of near-field sedimenting spray drift was demonstrated and quantified. A non-linear drift prediction equation was set up and validated, to predict the expected magnitude of drift for the reference spray application as a function of drift distance, average wind speed at a height of 3.25 m, average temperature and absolute humidity. This equation shows that decreasing wind speed and temperature and increasing absolute humidity decreases the amount of sedimenting spray drift and stresses the important effect of air humidity and temperature. This equation was used to compare the drift results of the different spraying techniques under various weather conditions with the reference spraying by calculating their drift reduction potential (DRP_t).

In total, 162 PDPA laser measurements, 51 wind tunnel experiments and 108 field drift experiments were performed. From these measurements, droplet size and velocity characteristics at a nozzle distance of 0.50 m, drift potential reduction percentages (DPRP_V1, DPRP_V2 and DPRP_H) and drift reduction potentials (DRP_t) were determined and compared for the different spray application techniques to investigate their effect on the amount of near-field sedimenting spray drift and to evaluate the potential of the different drift assessment means.

From the PDPA measurements, it was found that droplet sizes vary from a few up to some hundreds of micrometres and droplet velocities from about 0 m.s^-1 up to 16 m.s^-1. Droplet sizes and velocities are related and both are influenced by nozzle type, size and spray pressure. Droplet velocities at 0.50 m are mainly determined by the ejection velocity and by their size. Smaller droplets slow down more rapidly due to the effect of air drag compared to larger droplets. Although bigger droplet sizes generally correspond with higher droplet velocities (and vice versa) droplet velocities for one and the same droplet size range vary depending on nozzle type and size because of variations in ejection velocities. The PDPA measuring set-up was capable of producing huge amounts of useful and repeatable droplet velocity and size information under controlled conditions. Comparing the PDPA measuring results with other studies confirmed the need for reference nozzles to classify sprays because of the considerable variation of absolute results depending on measuring protocol, settings, type and variations in reference sprays and measuring equipment.

Standard flat fan nozzles produced the finest droplet size spectrum followed by low-drift and the air injection nozzles which results in significant differences in the proportion of small droplets (e.g. V₁₀₀, V₂₀₀) and different other droplet size characteristics like D_v0.5, RSF, D₁₀, etc. The effect of nozzle type on droplet sizes was more important for the smaller ISO nozzle sizes. For the same droplet size, velocities are the highest for the flat fan nozzles followed by the low-drift and the air inclusion nozzles which is caused by differences in ejection velocities resulting from the pre-orifice effect in case of a low-drift nozzle and by a combination of venturi and pre-orifice effect for the air inclusion nozzles. In spite of this, droplet velocities are generally highest for the air inclusion nozzles, followed by the low-drift and the standard flat fan nozzles - for the same ISO nozzle size and spray pressure - because of their coarser droplets corresponding with higher velocities. Hence, the droplet size effect dominates the ejection velocity effect. From the wind tunnel and field experiments, highest DRPt and DPRP values are found for the air inclusion nozzles followed by the low-drift and the standard flat fan nozzles. Again, the effect of nozzle type is most important for the smaller nozzle sizes.

The larger the ISO nozzle size, the coarser is the droplet size spectrum and the lower is the proportion of small droplets. This effect is most pronounced for the standard flat fan followed by the low-drift nozzles and is less important for the air inclusion nozzles. Bigger ISO nozzle sizes also correspond with higher droplet velocities for the same nozzle type and spray pressure. This is caused by two factors which strengthen each other namely, bigger ISO nozzles produce bigger droplets which are faster and droplets of the same size produced by bigger nozzles are faster because of higher ejection velocities. These effects of nozzle size on droplet characteristics are clearly reflected in the results from the wind tunnel and the field measurements. The bigger the ISO nozzle size, the higher the DPRP and DRP_t values for the standard and the low-drift flat fan nozzles. For the air inclusion nozzles, the effect of nozzle size on DPRP and DRP_t values is less clear but in all cases, DPRP and DRP_t values are high and the highest values are found for the ISO 03 nozzles.

To investigate the effect of spray pressure, a series of measurements was carried out with the ISO 03 standard flat fan nozzle within a pressure range from 2.0 to 4.0 bar. For the droplet velocities, only the fastest droplet velocity characteristics (vvol75 and vvol90), significantly decrease with decreasing pressures. Although decreasing pressure from 3.0 to 2.0 bar did not significantly affect droplet size characteristics, fallout and airborne downwind spray deposits in the wind tunnel significantly increased because of this slight reduction in droplet velocities in combination with a decrease in entrained air velocities. On the other hand, this decrease in spray pressure resulted in a clear decrease in the amounts of field drift which was in contrast with the results from the wind tunnel and the PDPA laser measurements. Increasing the spray pressure from 3.0 to 4.0 bar significantly decreased droplet sizes but the effect was limited compared to the effect of nozzle size. In the field, an increase in spray drift was found.

Besides nozzle type, size and spray pressure, all having an effect on spray quality, driving speed and boom height also influence the amount of spray drift. Based on the field and the wind tunnel experiments, it was found that operating at a boom height as close as possible to the vegetation - without sacrificing the spray pattern - is a good way to reduce drift. The effect of driving speed could only be investigated in a realistic way in the field. A decrease in spray drift is observed for lower driving speeds of 4 and 6 km.h_-1 while the difference between the reference speed of 8 km.h_-1 and a speed of 10 km.h_-1 is statistically non-significant.

Looking at the effect of air assistance, a reducing effect on the total amount of spray drift is demonstrated for the Hardi ISO F 110 02, F 110 03 and LD 110 02 nozzles with drift reduction factors a_d of, respectively, 2.08, 1.77 and 1.53. No significant effect was found for the LD 110 03 nozzles which demonstrates that the finer the spray, the higher the impact of air assistance is on the amount of spray drift.

Comparing the results of the three drift assessment means, droplet size as well as velocity characteristics are related with field measurement DRP_t values and wind tunnel DPRP values. In general, DRP_t and DPRP values increase with increasing values of droplet diameter and velocity characteristics and decrease with increasing percentages of small droplets.

The proportion of the total volume of droplets smaller than 200 µm (V₂₀₀), was found to be the best individual indicator for the amount of sedimenting spray drift with an R² of 0.90. Besides V₂₀₀, the droplet size characteristics V₅₀, V₇₅, V₁₀₀, V₁₅₀ and V₂₅₀ and the velocity span factor (VSF) were also strongly related with DRP_t. The higher the VSF value, representing a less uniform droplet velocity distribution, the lower the DRP_t value which can be explained by the clear relation between droplet sizes and velocities which is reflected in the VSF values.
With regard to the wind tunnel measurements, the different individual droplet size characteristics are best related with DPRP_H followed by DPRP_V2 and DPRP_V1, the opposite is found for the droplet velocity characteristics. With regard to DPRP_H, V₁₀₀, V₁₅₀ and V₂₀₀ have the highest predictive power (R² = 0.92), while DPRP_V1 was related most with vvol10 (R² = 0.86) and DPRP_V2 with VSF (R² = 0.90) which shows again that droplet sizes and velocities are linked and that droplet size characteristics are more related with fallout compared to airborne deposits while the opposite is found for the droplet velocity characteristics.

A fairly good correlation was found between field drift DRP_t and wind tunnel DPRP values with the best agreement with DPRP_H (R² = 0.88) followed by DPRP_V2 (R² =0.81) and DPRP_V1 (R² = 0.66). Moreover, similar trends are found - concerning the effect of nozzle type, size, height and pressure - from the DPRP and DRP_t results although there are some deviations in absolute results mainly for a varying spray pressure and nozzle height. This means on the one hand that the wind tunnel approach calculating the surface under the fallout deposit curve, is best suited to represent real near-field sedimenting drift characteristics. On the other hand it indicates that the indirect drift assessment method measuring V₂₀0 values is at least as well suited to represent near-field drift characteristics compared with the wind tunnel approach calculating DPRP_H and even better suited compared with the wind tunnel approaches calculating DPRP_V1 and DPRP_V2. With the PDPA laser, it is only possible to investigate the effect of nozzle type, size and spray pressure whereas the effect of nozzle height can also be investigated by means of wind tunnel measurements. With both indirect techniques, it is difficult to investigate effects like driving speed and air assistance where direct drift measurements are necessary. Field research is appropriate for obtaining realistic estimates of drift under a range of working conditions but it is time-consuming and expensive. In this study, a measuring protocol and a drift prediction equation were set up to improve the interpretation of field drift data. With this equation and DRP_t of a certain spray application technique, realistic sedimenting field drift data for varying meteorological conditions can be calculated. With the indirect drift assessment means, driftability experiments can be made with different spraying systems under directly comparable and repeatable conditions and both methods are suited to permit relative studies of drift risk. Moreover, based on DPRP_H or V₂₀₀ - resulting from wind tunnel and the PDPA measurements - the DRP_t of a certain technique can be determined to come to a realistic estimate of field drift data at a driving speed of 8 km.h^-1 and a boom height of 0.50 m. This information is useful for all users of plant protection products, constructors and authorities for decision-making and risk assessment processes.

With this study, a large database with droplet characteristics, wind tunnel fallout and airborne deposits and (absolute) near-field drift results of different spray application techniques is made available together with information about the effects of climatological conditions. The results of this research are in fairly good agreement with the results from different other studies although it is difficult to compare because of differences in, among others, spray application techniques, experimental design, tracers and weather and crop conditions. That is why it is increasingly important to unify the different indirect and direct drift assessment means and to put together the different available databases.